Starting from a randomly generated state of the 15-puzzle game (https://en.wikipedia.org/wiki/15_puzzle), steepest-ascent hill-climbing (the vanilla version of hill-climbing search) gets stuck 76% of the time, i.e., solving only 24% of problem instances. But it works very quickly, i.e., it takes just 6 steps on average when it succeeds and 5 steps when it gets stuck. In contrast, if sideways moves are allowed, this raises the percentage of problem instances solved by hill-climbing from 24% to 81%, with the success at a cost: the algorithm averages roughly 7 steps for each successful instance and 32 steps for each failure. Now suppose that we are implementing random-restart hill climbing (i.e., if a search fails, it keeps to try, and try, until it gets a success) by the following two versions: one uses vanilla steepest-ascent hill climbing, and the other one uses hill climbing with sideways moves. Can you please tell which version of random-restart hill-climbing listed above runs faster on average? Please justify your answer.
Starting from a randomly generated state of the 15-puzzle game (https://en.wikipedia.org/wiki/15_puzzle), steepest-ascent hill-climbing (the vanilla version of hill-climbing search) gets stuck 76% of the time, i.e., solving only 24% of problem instances. But it works very quickly, i.e., it takes just 6 steps on average when it succeeds and 5 steps when it gets stuck. In contrast, if sideways moves are allowed, this raises the percentage of problem instances solved by hill-climbing from 24% to 81%, with the success at a cost: the
- one uses vanilla steepest-ascent hill climbing, and
- the other one uses hill climbing with sideways moves.
Can you please tell which version of random-restart hill-climbing listed above runs faster on average? Please justify your answer.
Trending now
This is a popular solution!
Step by step
Solved in 2 steps
